A study on chloride induced depassivation of Fe-P-C-Si and Fe-P-C-Si-N steels in simulated concrete pore solution

The corrosion behaviour of high phosphorous steels containing varying amounts of silicon and nitrogen was studied by potentiodynamic polarization, linear polarization resistance (LPR) and electrochemical impedance spectroscopy (EIS) measurements. The morphology of a steel specimen tested in chloride containing concrete pore solution was studied using scanning electron microscope (SEM) and the elemental distribution at the pitting corrosion area was investigated using electron dispersive spectroscopy (EDS). The results showed that the capacitance increased and resistance declined with immersion time in Ca(OH)2 solution containing 0.1% chloride for plain carbon steel. The opposite was observed in the case of the high phosphorous steels. The potentiodynamic polarization and LPR results complement the EIS findings. Corrosion behaviour could be described with an equivalent circuit having two time constants. The creation, expansion and degradation of the passive layer were discussed with the help of the equivalent circuit elements. The SEM-EDS studies revealed that MnS inclusions at the surface could have a role in the initiation and growth of pits and that phosphorous was present at the pit free surface of the steel.


Introduction
The corrosion resistance of the Gupta era Delhi Iron Pillar that resisted corrosion for 1600 years is attributed to its phosphorous content [1] [2]. P-rich iron beams used in Puri and Konark withstood marine coastal environments for hundreds of years [3]. However, phosphorous is known to cause cold shortness in cast and wrought steel because of its tendency for grain boundary segregation and therefore its content is restricted to <0.05% [4] [5]. Ductile phosphoric irons can be produced by adding a small amount of carbon (0.01 wt % -0.05 wt %) to the steel [6] [7][8] [9]. Ductility can be further improved by applying duplex phase heat treatments which remove phosphorous from the grain boundaries improving cohesion [10]. Grabke and Erhart reported that nitrogen can expel phosphorous from the grain boundary sites and decrease the percentage of intergranular fracture [11]. An impact energy value of 118±7.8 J is reported for a Fe-0.15%P steel [12]. Liu and co-workers [13] found that adding 0.5% Si to high phosphorous iron alloys reduced the ductile-brittle transition temperature (small scale Charpy impact test) by approximately 100 K, and the fracture mode became transgranular. The corrosion resistance of Fe-P steels was observed to be better than that of currently used thermo-mechanically treated bars in chloride-containing simulated concrete environments [14].
Concrete contains pores which hold hydrated cement paste. Concrete always contains moisture unless it is dried in an oven [15]. Ions in the paste dissolve in the pore moisture and the resulting solution has a pH of about 12.5. This ensures that the reinforcement bar is in the passive state owing to the creation of a protective passive layer on the surface of the embedded rebar steel [15] [16]. This passivity may be destroyed in cases of leaching out of fluids from concrete, atmospheric carbonation or uptake of chloride ions [17] [18]. Chloride ions induce corrosion of passive reinforcing steel especially when deicing salts or chloride contaminated aggregate are incorporated into the concrete [17] [19]. Marine exposure is also a main source of Cl − ions.The presence of chloride ions above a threshold value near the rebar surface inside concrete depassivates the steel [20][21] [22] [23]. Ingress of atmospheric carbon dioxide converts calcium hydroxide to calcium carbonate and reduces pH promoting corrosion [19]. The breakdown of passivity leads to corrosion of the rebars.
Silicon additions to steel favor the formation of goethite with smaller particle size [24]. Nitrogen improves the resistance to localized corrosion like pitting and crevice corrosion in steel [25] [26]. The Fe-P-C-Si and Fe-P-C-Si-N steels developed in the present work have impact energy values of about 230 J at room temperature in the heat treated condition and are promising in concrete reinforcement (Table 1 and 3). Therefore it is important to study the corrosion behavior of these steels in Ca(OH)2 solution. Corrosion of steel in concrete can be evaluated by EIS as it is a non-destructive (AC< 20mV), effective in-situ monitoring method and suitable for obtaining the kinetic information related to corrosion reactions occurring on the steel surface [22] [27].
Therefore, in the present work the corrosion behavior of novel Fe-P-C-Si and Fe-P-C-Si-N steels in simulated concrete pore solution with/without various amounts of chloride ions is investigated. Plain carbon steel is included for comparison. The behaviour is studied using EIS. Further, potentiodynamic polarization curves and linear polarization resistance measurements are employed to understand the corrosion behaviour. The mechanism of formation and degradation of the passive film on these novel steels is explained on the basis of the results of the aforesaid methods.

Experimental procedure 2.1 Preparing the alloys
Three high phosphorous steels were manufactured using an induction furnace. Desired compositions were produced from melting scrap iron, graphite, ferro-silicon, nitrided manganese, and ferro phosphorous. Aluminium shots were added for de-oxidation. Plain low carbon steel was included in the study for comparison purposes. The chemical composition of the four steels obtained using Magellan optical emission spectrometer is shown in table 1. The mean composition after two readings is presented in the table 1.

Corrosion behaviour
The corrosion study sample (15 mm x 15 mm x 8 mm) surfaces were polished with emery paper (up to 800 grit) and cleaned with acetone. Three samples were prepared for ensuring reproducibility in experimentation. A standard flat cell (Ametek USA) was used for testing the samples. Potentiodynamic polarization, linear polarization resistance and electrochemical impedance spectroscopy studies were conducted. Potentiodynamic polarization experiments were run using an EG&G PARSTAT 273A Potentiostat (Ametek, USA) at a scan rate of 0.75 mVs −1 with SCE reference electrode. Saturated Ca(OH)2 solutions with 0, 0.05, 0.1, 0.14 weight % Cl − Content were employed (All future references to % Cl − in this text should be construed to be weight %). Earlier work has shown that results of electrochemical tests conducted in saturated Ca(OH)2 are similar to those obtained in concrete [28]. The pH of the solution (12.5) was measured using pH 510 of Eutech instruments. Before commencing the potentiodynamic polarization experiments, the open-circuit potential (OCP) was allowed to stabilize for 1h.
The EIS and LPR experiments were conducted using Gamry 1000 Interface at intervals of 15 minutes, 5h and 21h. For LPR, the potential was scanned at 0.125 mV s −1 from −20mV to +20 mV with respect to the OCP. The slope of the E vs i curve at zero current yielded the polarization resistance (Rp). LPR technique was chosen to monitor the corrosion rate during immersion of samples for 21 h since the surface of the sample does not deteriorate. LPR was useful in studying the change in the corrosion resistance of the passive surface with time in the test solution without and with chloride. The measured LPR can be used to calculate the instantaneous corrosion rate using the Stern-Geary equation [29][30] as shown below: Z depends upon the anodic (βa,) and cathodic (βc) Tafel slopes. For each solution composition, the test was repeated twice. For the EIS tests, a sinusoidal potential perturbation of 10 mV was applied at the OCPs with frequency sweep from 100 KHz to 10 mHz using the GAMRY Interface 1000 potentiostat. GAMRY EChem analyst software was used to analyse the EIS data by fitting with equivalent circuits. These analyses determined the passive layer parameters. The high frequency arc was fully revealed by scanning from 2MHz to 1000 Hz using PARSTAT 2273. The solution resistance and the adsorbed layer data was calculated using this data.
The surface morphology of S3 steel specimen used for potentiodynamic polarization in the Ca(OH)2 solution containing 0.14% chloride was observed by SEM (Zeiss EVO 18) in order to examine its state of corrosion. Elemental distribution at pitted area was studied using EDS.

Dilatometry and mechanical behaviour
The results of the dilatometry studies are shown in table 2. These were used to select the temperature of 900°C for the duplex phase heat treatment of the tensile and impact test samples. The mechanical properties of these steels hot forged and subsequently heat treated for 1 hour at 900°C condition are shown in table 3. The strength of all the high phosphorous steels is around 400 MPa and the elongation exceeds 33%. The yield stress, ultimate tensile stress, % elongation and % reduction in area of S1, S2 and S3 steels compare well with the findings of other researchers for Fe-0.11P-0.03C steel [10]. The toughness of S1, S2 and S3 steels is better than the energy value of 118±7.8 J recorded for Fe-0.09C-0.37Si-0.42Mn-0.15P-0.37Cr-0.35Cu-0.31Ni-0.03Al steel [12]. The impact values of the high phosphorous steels studied in the present work are comparable with those of the plain carbon steel. Encouraged by these good mechanical properties, the corrosion behaviour of the high phosphorous steels is studied and discussed in the following sections.   Figure 1 shows the anodic polarization curves of the processed steel samples tested in saturated Ca(OH)2 solution containing various Cl − ion concentrations. All the samples exhibit passive behaviour in this solution. This conforms to the Fe-water Eh-pH diagram of 298.15 K. The range of passivity decreases as the Cl − content increases beyond a threshold value. In all the curves for all the samples immersed in solution with 0.05% Cl − and without chloride, current density increases at about 650 mVSCE which is attributable to O2 evolution and is dependent on pH. Beyond a critical chloride concentration, the corrosion current registered a drastic increase during the experiment preceding the oxygen evolution. In the presence of Cl − ions, the pitting potential became more negative towards the active side. For each steel composition, the pitting or passive film breakdown occurred when the chloride concentration exceeded a critical amount. The threshold value of Cl − ion concentration for the initiation of pitting is determined as 0.1% Cl − for S1, S2 and S3 steels while it was 0.06% Cl − for the S4 steel. At these concentrations the pitting potential of the steel samples was lower than the oxygen evolution potential.

Anodic polarization behavior
Tests were also carried out in solutions containing 0.14 Cl − in order to confirm the lower pitting potential of the steels. Other researchers have reported threshold values of 0.05% Cl − for AISI 1010 steel [31], 0.06% Cl − for Fe-0.15C steel and 0.1% Cl − for Fe-0.11P steel [14]. Thus, S1, S2 and S3 steels are more resistant to pitting than S4 steel as their threshold chloride is found to be higher in this work.

OCP and LPR studies
For all steel samples, the open circuit potential and the linear polarization resistance were recorded after 15 minutes, 5 h and 21 h. The LPR values were used to calculate icorr using equation (1). Figure 2 shows the variation of the OCP and icorr for the samples S3 and S4 steels exposed to aerated saturated solution of Ca(OH)2 containing various chloride concentrations. The OCP of S3 steel shifted towards noble direction while the icorr values decreased with time of dipping in all solutions. Similar trends were observed in the case of S1 and S2 steels too. This trend implies that with increasing immersion time the steel became nobler [32]. It has been reported that as the anodic curve shifts to lesser current densities, the Ecorr increases and icorr decreases [33]. In the case of S4 steel immersed in solutions containing 0.05% and 0.1 % Cl − , there is an overall decline in OCP with the immersion time which suggests increased corrosion. The icorr values of S4 steel did not show a corresponding overall increase. The OCP values indicate that the critical chloride concentration for initiation of pitting could be 0.05% and is probably lower than that of the high phosphorous steel samples.  Figure 2. Variation of the OCP and icorr for the S3 and S4 steel samples exposed to simulated concrete pore solution containing various chloride concentrations. Figure 3 shows the Nyquist, Bode magnitude and Bode phase plots of S3 and S4 steels in saturated solution of Ca(OH)2 containing 0.1% Cl − at different times of immersion. The impedance responses of S3 steel in saturated solution of Ca(OH)2 without chloride and that having 0.05% Cl − were similar to those obtained in solution having 0.1% Cl − . For S1, S2 and S3 steels, the impedance escalated with time in all solutions. This variation in response was also found similar to that of S4 steel in Ca(OH)2 solution containing zero and 0.05% chloride. A different impedance response was recorded for S4 steel in Ca(OH)2 solution with 0.1% chloride concentration. A capacitive loop of depressed semicircles is seen in the Nyquist plots of S3 steel. Further, with increase in immersion time the capacitive part increases. Again, with escalation in immersion time, the polarization resistance or the real impedance increases. Opposite trend is seen in the Nyquist plots of S4 steel i.e. as the time of immersion increases there is a decline in the polarization resistance and capacitive component. Upon dipping a metal in a solution, the metal ions enter the solution and in time an electrical double layer is formed at the metal-solution interface. The current at the metal electrolyte interface can be modelled as a capacitance as C=q/V. Here, q represents the charge in the double layer upon dissolution of the metal. V is the potential at which the sample corrodes freely. If the passive film grows becomes more protective with time then the micro-pores present on the film will grow smaller with time and metal dissolution will decrease. This implies that the capacitance associated with the double layer will decrease since C=1/2πfZ where f is the frequency. The capacitive part increased and capacitance decreased with time of immersion for S1, S2 and S3 steels. Again

Electrochemical impedance spectroscopy
where d is the thickness of the film, , ε0 is the vaccum permittivity and ε is the dielectric constant of the film [34]. Therefore, a thicker and more protective passive layer with time of immersion implies a decrease in capacitance C as observed in the case of S1, S2 and S3 steels. The capacitance increased and capacitive part decreased with immersion time, due to an increased q owing to corrosion of S4 steel in (Ca(OH)2 + 0.1 Cl − ) solution. Real impedance decreased with immersion time indicating breakdown of passive film.
For S3 steel tested in the saturated solution of Ca(OH)2 without Cl − and with 0.05 % Cl − and 0.1% Cl − , at frequencies less than about 10 2 Hz the Bode magnitude plots exhibited a purely capacitive behaviour. With increasing immersion time, the straight-line portion with a slope of about −1 (the capacitive part) expanded. This response was found similar to the responses of S1 and S2 steels. Similar response was observed for S4 steel in Ca(OH)2 solution and (Ca(OH)2 + 0.05% Cl − ) solution, too. At the end of 21 h of immersion, the impedance of passive film ranged from 10 4 Ω.cm 2 to 10 5 Ω.cm 2 for all the samples. The passive film acts as a capacitor effectively when the impedance is greater than 10 6 Ω.cm 2 [33]. The high values of impedance imply that the passive film becomes more protective with time of immersion. For S4 steel in (Ca(OH)2 + 0.1 Cl − ) solution, the central portion of bode magnitude curve declined with time of immersion, which indicates that the passive film broke down.
By measuring the high and low break point frequencies of the Bode magnitude plots, capacitance can be calculated. These frequencies are found where the two frequency independent horizontal parts intersect with the central straight line portion having negative slope [35]. The low and high frequencies can be located at the 45° phase angle too. The capacitance corresponding to high (fh) and low (fl) frequencies can be calculated as follows: Where Ru is the solution resistance and Rp is the polarization resistance. Cl declined with time indicating enlargement of passive layer for S1, S2 and S3 steels in all the solutions and S4 steel in solutions containing less than 0.1 Cl -. The fl of S4 steel for the (Ca(OH)2 + 0.1 Cl − ) solution increased with immersion time depicting the breakdown of passive layer and dissolution of iron. Capacitance is considered to be a good pointer for the initiation of pitting [33] [36].
In the Bode phase plots, the phase angle (θ) in S1, S2 and S3 steels in all solutions and S4 steel in solutions containing less than 0.1 Cl − declined to zero at fh and fl. The value of θ at zero correlates with the resistances of Ru and (Ru+Rc+Rp). Capacitive behavior is indicated by the increase of θ towards 90° at intermediate range of frequencies [35]. The crest of Bode phase plots in S1, S2 and S3 steels in all solutions and S4 steel in solutions containing less than 0.1 Cl − widened with time of immersion. This considerable phase shift at absolute value of θ (crest of the plot) in the Bode phase angle graph implies broadening of passive layer with immersion time. The frequency, at which the decline in absolute value of θ starts, decreased as the immersion time increased. As the immersion time increased, the lower frequency break point corresponding to 45° drifted towards lower frequencies. Therefore, the phase angle drift at low frequencies increased with immersion time for Bode phase plots in S1, S2 and S3 steels in all solutions and S4 steel in solutions containing less than 0.1 Cl − . This indicates improvement of protective capabilities of the passive layer with immersion time [33][36] [14]. The degradation of organic coatings are indicated by a decrease in phase angle shift at low frequency regions [14] [36]. The phase angle drift at low frequencies decreased with immersion time for Bode phase plots in S4 steel in solutions having 0.1 Cl − indicating degradation of passive layer and increasing corrosion with immersion time.
From the above observations on the Nyquist and Bode plots one can say that the decreasing capacitance and increasing impedance was found in the cases S1, S2 and S3 steels in all solutions and in case of S4 steel in solutions containing less than 0.1 Cl − . This may be due to decrease in the micropores and increased protective coverage of the surface by the passive oxide layer with immersion time. The film deteriorated in the case of S4 steel in solutions having 0.1 Cl − and corrosion increased with time. Thus it can be said that for S1, S2 and S3 steels, the threshold Cl − concentration for degradation of oxide layer is higher than 0.1%. It is less than 0.05% for S4 steel.   At frequencies lower than 790 Hz, the arc which appears is related to the interfacial reaction. This reaction is composed of two processes viz. the charge transfer reaction and the formation of a passive film, which have an identical time constant as per the opinion of other researchers [17] [37]. In order to fit the experimental data, many equivalent circuits were tried using Gamry Echem Analyst software version 6.04. The best fitting equivalent circuit that approximates the test system is shown in figure 5. It can also be written as Ru (Cc(Rc(Y0Rp))). Here, solution resistance is represented by Ru. The resistance and capacitance of the adsorbed layer are given by Rc and Cc. Rp and Y0 are the resistance and capacitance associated with the corrosion processes occurring in the micro-pores of the oxide film. The oxide film resistance is composed of the resistance due to the film and charge transfer. Mathematically, the impedance of the circuit can be written as where ω is the angular frequency and  Very little variation was found in the parameters Ru, Rc and Cc with respect to immersion time and chloride content. In the simulated concrete pore solution either with or without chloride, mean Ru was about 17 Ω.cm 2 , mean Rc was about 80 Ω.cm 2 and mean Cc was about 0.019 μF/cm 2 for all the samples (Figure 4). Ru, Rc and Cc for mild steel and Fe-P alloys in Ca(OH)2 solution with/without Cl − found by Sahoo and Balasubramaniam were approximately 25 Ω.cm 2 , 35 Ω.cm 2 and 11 μF/cm 2 respectively [14]. Rs, Rc and mean Cc for mild steel in simulated concrete pore solution with/without Cl − were found by Saremi and Mahallati to be 62 Ω.cm 2 , 31 Ω.cm 2 and 0.02 μF/cm 2 respectively [17]. Saremi and coworkers attributed the adsorbed layer to OH − ion on the surface of iron. The miniscule resistance of this adsorbed layer does not help in making the steel corrosion resistant.
The variables concerned with the passive film polarization resistance viz. Rp, CPE exponent α and constant phase element Y0 showed variation. The double layer-capacitance impedance can be expressed by a constant phase element (CPE) rather than an ideal capacitor in case the surface is rough and heterogeneous [14][22] [38][39] [40]. The impedance (ZCPE) and capacitance (CCPE) are defined by equations (7) and (8) Y0 represents a combination of properties related to electroactive species and the steel surface. It is proportional to the capacitance of pure capacitive electrodes [14] [22]. α is the CPE exponent. It is a phenomenological coefficient which indicates the degree of deviation in the capacitance from the ideal condition [22][42] [43]. The frequency corresponding to maximum phase angle is ωm [22]. Y0 is called CPE when α lies between 0.5 and 1 and is based on surface heterogeneity and roughness [14]. Y0 is pure capacitance when α =1 and pure resistance when α =0 [44]. Rp values generally escalated with immersion time for S1, S2 and S3 steels in all solutions and S4 steel in solutions containing less than 0.1 Cl − . Rp values decreased with time for S4 steel in solutions containing 0.1 Cl − (Figure 6a). Rp values of all the samples vary between 1.4x10 4 to 7.7x10 5 Ω.cm 2 . The Rp values obtained by Sahoo and Balasubramaniam after 7 days immersion of Fe-P alloys and mild steel in Ca(OH)2 solution with/without Clvaried between 1.5x10 4 and 7.3x10 5 [14]. The Rp values for plain carbon steel in Ca(OH)2 solution without and with 0.3 Clwere found by Saremi and Mahallati to be between 1.5x10 5 to 1.7x10 5 Ω.cm 2 [17]. Reported values for passive layer of steel in borate buffer alkaline solution are 10 4 to 10 6 Ω.cm 2 [45] [46].

Rp
Y0 values were between 36 to 77 μS s α /cm 2 which agrees with the values of 14-75 μS s α /cm 2 for phosphoric irons and mild steel found by Sahoo and co-workers [14]. An overall decrease in Y0 was observed with immersion time due to the expansion of passive layer except in the case of S4 steel in Ca(OH)2 solution containing 0.1% chloride (Figure 6b). This is caused due to breakdown of passive film and consequent increase in capacitance [14] [17]. α values varied between 0.78 to 0.92 which confirms the heterogeneity of the steel surface as observed by other researchers too [38] [39]. The increase in α of S3 steel indicates improvement in passive film and the decrease in α of S4 steel implies that the film became increasingly inhomogeneous and defective with time ( Figure 6c) [22]. Figure 7 shows a schematic diagram of the growth or destruction of the passive layer with time. On immersion in the solution the passive film starts forming and iron ions enter the solution after passing  (Figure 7b) they become smaller reducing the exposed area and hence the corrosion current. This increase in thickness of double layer causes an increase in Rp and decrease in capacitance (follows decrease in charge per unit area). In the case of S4 steel immersed in simulated concrete pore solution containing 0.1% Cl − , the chloride ions destabilize the passive film by migrating through the micro-pores and probably causing the following reaction on the metal surface exposed inside the pores [14]. 2+ + 2 2 + 2 − → ( ) 2 + 2 (9) The Cl − formed may be used in auto-catalysis of the pitting process, thereby increasing the size of the micro-pores and leading to degradation of passive layer and accelerating the corrosion processes ( Figure  7d).

Corrosion surface analysis
Surface morphology of S3 steel after potentiodynamic polarization testing in (Ca(OH)2 + 0.14% Cl − ) solution was observed by SEM to examine its state of corrosion. Surface shows a few pits and sporadic deposits. The observations agree with the outcome of the electrochemical studies. EDS was employed to understand the mechanism of corrosion of S3 steel from chlorides by recording the elemental distribution at the pitted location ( Figure 8). EDS reveals that Mn, S, Cl, O and Si are present in the pit region. This suggests that adsorption of chloride ions followed by the initiation of pits could have been favoured by presence of MnS inclusions. Literature suggests that chloride is adsorbed preferentially at the MnS inclusions associated with oxides of aluminium causing their dissolution accompanied with a local drop in pH leading to local autocatalysis [47]. Once a pit nucleates it might grow depending on whether the local chloride concentration exceeds critical limits [22]. Further, P, Cr and N are found welldistributed and deficient in the pit region. It can be stated that P could play a positive role in protecting the surface from pitting damage.

Beneficial effect of phosphorous alloying
The OCP measured vs SCE for all the samples are above −0.5 V. Therefore, according to the Eh-pH diagram of phosphorous-water system, phosphorous in the S1, S2 and S3 steels will remain in equilibrium with PO  3 4 in saturated Ca(OH)2 solution having pH 12.5, at and above −0.258 V vs SHE [48].
Anodic corrosion inhibitors employed for iron in concrete environments involve phosphates like CaHPO4, Na3PO4 and Na2HPO4 [28] [49]. The phosphate ions help in the oxidation of ferrous to ferric form. The drift of the ferrous ions away from the metal surface is prevented [50]. The accumulated ferric ions form a protective γ-Fe2O3 film in Ca(OH)2 solution [28]. The film causes a reduction in the intensity of corrosion by reducing active sites. Phosphorous containing steels form a bipolar passive film. The inner iron oxide layer is anion selective while the outer Fe3(PO4)2 enriched layer is cation selective and hence blocks the entry of Cl − ions into the film improving resistance against nucleation and growth of pits [51] [52].The inner layer probably converts to gamma Fe2O3 leading to the thickening of the passive layer [14]. This also explains the absence of P in the pit region and presence of P outside the pit region on the surface of the steel in figure 8.

Conclusions
The high phosphorous steels developed in this work showed mechanical properties comparable to rebar steel. Their polarization studies in simulated concrete pore solution containing up to 0.1% chloride concentration revealed a stable passive behaviour and good pitting corrosion resistance. Whereas, plain carbon steel could withstand, only up to 0.06% chloride. LPR and OCP measurements also confirmed the improved performance of phosphorous steel over the plain carbon steel. High capacitance from the equivalent circuit obtained from the EIS spectra indicated rough surface due to pitting corrosion. Decreased resistance of passive film implies that chloride above the threshold concentration degraded the passive layer of steel. With the thickening of the passive layer, phase angle plots widened. Unlike phosphorous steels, the plain carbon steel corroded actively in 0.1% chloridecontaining simulated concrete pore solution.
Corroded surface of S3 steel tested in simulated concrete pore solution with 0.14% chloride solution suggest that MnS inclusions could be crucial in the initiation of pits since they attract chloride ions. The presence of phosphorous outside the pits could indicate that phosphorous blocks the entry of chloride ions and thus prevents corrosion. Thus, the high phosphorous steels developed in this work have potential to deliver improved corrosion performance in concrete environments provided weldability concerns are suitably addressed.